Abstracts

Rationale: There are serious morbidities associated with nonepileptic psychogenic status (NEPS). Whether patients with markedly prolonged psychogenic nonepileptic seizures (PNES) or NEPS are distinguishable from patients with shorter duration PNES is not known and would be important for reducing intubation risk and other iatrogenic complications. Methods: We compared two groups of patients with video EEG (vEEG) confirmed PNES based on event duration. Seventy patients were consented and prospectively enrolled in a PNES study and given a semi-structured interview and multiple questionnaires including one that asked for the duration of the longest event ever experienced. Twenty five of 70 reported that they never had any event lasting more than 5 minutes and were considered the ""short PNES"" group. We retrospectively identified 36 patients from our EMU database with vEEG-captured markedly prolonged PNES ranging from 10-45 minutes; Fifteen of 36 had events lasting 20 minutes or longer and were considered as ""markedly prolonged PNES or NEPS"". Psychiatric and neurologic consultation notes, EEG, and imaging were available for these 15 patients and were reviewed, gathering information regarding demographics, event semiology, EEG and neuroimaging, and psychiatric comorbidity. This study was approved by the institutional review board (IRB). Results: There were no group differences in gender, ethnicity, education, living situation, employment, age of onset, duration of illness, or number of patients with comorbid epilepsy. The overwhelming majority were female (>80%) and white (>73 %) in both groups. Using a chi-squared test, NEPS patients were more likely to be admitted emergently through the ED or ICU transfer (23.1% vs. 4.1%, p=0.014) and less likely through a neurology office. NEPS were also more likely to involve convulsive movements (p=0.04), loss of consciousness (LOC) (p=0.038), and eye closure (p=0.027). Using a two-sample t-test, NEPS patients had significantly more semiologic features during events (p<.05) compared to those with short PNES. Groups did not differ with respect to EEG or MRI abnormalities (approximately 50% abnormal), or psychiatric comorbidity (including past suicide attempts and trauma history). As a secondary analysis, a stepwise logistic regression analysis was performed to evaluate clinical neurological predictors which identified total number of semiologic features during events as a significant predictor of NEPS vs. short duration PNES (odds ratio=4.80 for each additional feature; p=0.032). Conclusions: NEPS are more likely to have convulsive semiology and LOC as well as more varied semiology. Whether longer events with more semiologic features are a marker of more psychopathology has not yet been determined perhaps due to the greater likelihood of emergent over elective admission and the difficulty in capturing consistent prospective data.